Amanda M.Y. Chu , Yasuhiro Omori , Hing-yu So , Mike K.P. So
{"title":"A Multivariate Randomized Response Model for Sensitive Binary Data","authors":"Amanda M.Y. Chu , Yasuhiro Omori , Hing-yu So , Mike K.P. So","doi":"10.1016/j.ecosta.2022.01.003","DOIUrl":"https://doi.org/10.1016/j.ecosta.2022.01.003","url":null,"abstract":"<div><p><span>A new statistical method is proposed to combine the randomized response technique, probit modeling, and </span>Bayesian analysis<span> to analyze large-scale online surveys of multiple binary randomized responses. The proposed method is illustrated by analyzing sensitive dichotomous randomized responses on different types of drug administration error from nurses in a hospital cluster. A statistical challenge is that nurses’ true sensitive responses are unobservable because of a randomization scheme that protects their data privacy to answer the sensitive questions. Four main contributions of the paper are highlighted. The first is the construction of a generic statistical approach in modeling multivariate sensitive binary data collected from the randomized response technique. The second is studying the dependence of multivariate sensitive responses via statistical measures. The third is the calculation of an overall attitude score using sensitive responses. The last one is an illustration of the proposed statistical method for analyzing administration policies that potentially involve sensitive topics which are important to study but are not easily investigated via empirical studies. The particular healthcare example on drug administration policies demonstrated in this paper also presents a scientific way to elicit managerial strategies while protecting data privacy through analytics.</span></p></div>","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"27 ","pages":"Pages 16-35"},"PeriodicalIF":1.9,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50178657","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Estimation of Extreme Risk Measures for Stochastic Volatility Models with Long Memory and Heavy Tails","authors":"Clémonell Bilayi-Biakana, G. Ivanoff, Rafal Kulik","doi":"10.1016/j.ecosta.2023.07.004","DOIUrl":"https://doi.org/10.1016/j.ecosta.2023.07.004","url":null,"abstract":"","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"1 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76829796","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A Two-Way Transformed Factor Model for Matrix-Variate Time Series","authors":"Zhaoxing Gao , Ruey S. Tsay","doi":"10.1016/j.ecosta.2021.08.008","DOIUrl":"https://doi.org/10.1016/j.ecosta.2021.08.008","url":null,"abstract":"<div><p>A new framework is proposed for modeling high-dimensional matrix-variate time series via a two-way transformation, where the transformed data consist of a matrix-variate factor process, which is dynamically dependent, and three other blocks of white noises. For a given <span><math><mrow><msub><mi>p</mi><mn>1</mn></msub><mo>×</mo><msub><mi>p</mi><mn>2</mn></msub></mrow></math></span> matrix-variate time series, nonsingular transformations are sought to project the rows and columns onto another <span><math><msub><mi>p</mi><mn>1</mn></msub></math></span> and <span><math><msub><mi>p</mi><mn>2</mn></msub></math></span><span> directions according to the strength of the dynamical dependence of the series on their past values. Consequently, the data are nonsingular linear row and column transformations of dynamically dependent common factors and white noise idiosyncratic components. A common orthonormal projection method is proposed to estimate the front and back loading matrices of the matrix-variate factors. Under the setting that the largest eigenvalues of the covariance of the vectorized idiosyncratic term diverge for large </span><span><math><msub><mi>p</mi><mn>1</mn></msub></math></span> and <span><math><msub><mi>p</mi><mn>2</mn></msub></math></span><span><span>, a two-way projected Principal Component Analysis is introduced to estimate the associated loading matrices of the idiosyncratic terms to mitigate such diverging noise effects. A new white-noise testing procedure is proposed to estimate the dimension of the factor matrix. </span>Asymptotic properties of the proposed method are established for both fixed and diverging dimensions as the sample size increases to infinity. Simulated and real examples are used to assess the performance of the proposed method. Comparisons of the proposed method with some existing ones in the literature concerning the forecastability of the factors are studied and it is found that the proposed approach not only provides interpretable results, but also performs well in out-of-sample forecasting.</span></p></div>","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"27 ","pages":"Pages 83-101"},"PeriodicalIF":1.9,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1016/j.ecosta.2021.08.008","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50178660","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
V. Todorov, V. Simonacci, M. Gallo, Nikolay Trendafilov
{"title":"A novel estimation procedure for robust CANDECOMP/PARAFAC model fitting","authors":"V. Todorov, V. Simonacci, M. Gallo, Nikolay Trendafilov","doi":"10.1016/j.ecosta.2023.07.001","DOIUrl":"https://doi.org/10.1016/j.ecosta.2023.07.001","url":null,"abstract":"","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"47 1","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87478840","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Robust empirical risk minimization via Newton’s method","authors":"Eirini Ioannou, Muni Sreenivas Pydi, Po-Ling Loh","doi":"10.1016/j.ecosta.2023.07.003","DOIUrl":"https://doi.org/10.1016/j.ecosta.2023.07.003","url":null,"abstract":"A new variant of Newton’s method for empirical risk minimization is studied, where at each iteration of the optimization algorithm, the gradient and Hessian of the objective function are replaced by robust estimators taken from existing literature on robust mean estimation for multivariate data. After proving a general theorem about the convergence of successive iterates to a small ball around the population-level minimizer, consequences of the theory in generalized linear models are studied when data are generated from Huber’s epsilon-contamination model and/or heavy-tailed distributions. An algorithm for obtaining robust Newton directions based on the conjugate gradient method is also proposed, which may be more appropriate for high-dimensional settings, and conjectures about the convergence of the resulting algorithm are offered. Compared to robust gradient descent, the proposed algorithm enjoys the faster rates of convergence for successive iterates often achieved by second-order algorithms for convex problems, i.e., quadratic convergence in a neighborhood of the optimum, with a stepsize that may be chosen adaptively via backtracking linesearch.","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"260 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135568274","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Bayesian estimation for mode and anti-mode preserving circular distributions","authors":"Toshihiro Abe , Yoichi Miyata , Takayuki Shiohama","doi":"10.1016/j.ecosta.2021.03.004","DOIUrl":"https://doi.org/10.1016/j.ecosta.2021.03.004","url":null,"abstract":"<div><p><span><span>A Bayesian estimation is considered for unknown parameters of a unimodal skew circular distribution on the circle, where the underlying distribution has mode and anti-mode preserving properties. This distribution is obtained by using a transformation of the inverse </span>monotone function, and the shape of the resulting density can be flat-topped or sharply peaked at its mode. With regard to </span>Bayes estimates<span><span> (BEs), the boundary-avoiding priors are assumed so that the skewness and peakedness parameters of the distribution do not lie on the boundary of the parameter space. In addition to the BEs, </span>maximum likelihood estimations<span><span> (MLEs) are conducted to compare the performances in small samples, and found that the BEs are more robust than the method of maximum likelihood. As the pairs of parameters between location and skewness and between concentration and peakedness are independent of each other, approximate BEs using Lindley’s methods become rather simple. Monte Carlo simulations are performed to compare the accuracy of the BE and </span>MLE, and some circular datasets are analyzed for illustrative purposes.</span></span></p></div>","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"27 ","pages":"Pages 136-160"},"PeriodicalIF":1.9,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1016/j.ecosta.2021.03.004","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50178027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Bayesian analysis for mediation and moderation using g−priors.","authors":"Jean-Michel Galharret, Anne Philippe","doi":"10.1016/j.ecosta.2021.12.009","DOIUrl":"https://doi.org/10.1016/j.ecosta.2021.12.009","url":null,"abstract":"<div><p>A Bayesian analysis<span> is proposed using an extension of g-priors for moderated mediation models. For this choice of priors, an explicit form of the marginal distribution is obtained. Testing procedure on the existence of direct, indirect and moderated effects are constructed using Bayes factor approach. This methodology is applied to analyze the association between empowering leadership and organisational commitment in two companies.</span></p></div>","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"27 ","pages":"Pages 161-172"},"PeriodicalIF":1.9,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50178028","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Robust Covariance Matrix Estimation in Time Series: A Review","authors":"Masayuki Hirukawa","doi":"10.1016/j.ecosta.2021.12.001","DOIUrl":"https://doi.org/10.1016/j.ecosta.2021.12.001","url":null,"abstract":"<div><p><span><span>In the analysis of economic, financial and other time series, long-run variance estimators play an important role in estimating model parameters more efficiently and drawing more accurate </span>statistical inference on the parameters. A non-technical review of long-run variance estimation is provided. Both </span>parametric<span> and nonparametric estimators are discussed. Kernel methods are dominant among all estimation procedures, and therefore recent developments in kernel-smoothed estimators and related inference are presented. The information given can help practitioners decide on a suitable long-run variance estimator.</span></p></div>","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"27 ","pages":"Pages 36-61"},"PeriodicalIF":1.9,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50178656","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Forecasting value-at-risk and expected shortfall in large portfolios: A general dynamic factor model approach","authors":"Marc Hallin , Carlos Trucíos","doi":"10.1016/j.ecosta.2021.04.006","DOIUrl":"https://doi.org/10.1016/j.ecosta.2021.04.006","url":null,"abstract":"<div><p><span><span>Beyond their importance from the regulatory policy point of view, Value-at-Risk (VaR) and Expected Shortfall (ES) play an important role in risk management, portfolio allocation, capital level requirements, trading systems, and hedging strategies. However, due to the </span>curse of dimensionality<span>, their accurate estimation and forecast in large portfolios is quite a challenge. To tackle this problem, two procedures are proposed. The first one is based on a filtered historical simulation method in which high-dimensional conditional covariance matrices are estimated via a general </span></span>dynamic factor<span> model with infinite-dimensional factor space and conditionally heteroscedastic factors; the other one is based on a residual-based bootstrap scheme. The two procedures are applied to a panel with concentration ratio close to one. Backtesting and scoring results indicate that both VaR and ES are accurately estimated under both methods, which both outperform the existing alternatives.</span></p></div>","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"27 ","pages":"Pages 1-15"},"PeriodicalIF":1.9,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1016/j.ecosta.2021.04.006","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50178655","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Empirical Bayes Model Averaging with Influential Observations: Tuning Zellner’s g Prior for Predictive Robustness","authors":"Christopher M. Hans, Mario Peruggia, Junyan Wang","doi":"10.1016/j.ecosta.2021.12.003","DOIUrl":"https://doi.org/10.1016/j.ecosta.2021.12.003","url":null,"abstract":"<div><p><span><span>The behavior of Bayesian model averaging (BMA) for the normal linear regression model in the presence of influential observations that contribute to model misfit is investigated. Remedies to attenuate the potential negative impacts of such observations on inference and prediction are proposed. The methodology is motivated by the view that well-behaved residuals and good </span>predictive performance often go hand-in-hand. Focus is placed on regression models that use variants on Zellner's </span><span><math><mi>g</mi></math></span> prior. Studying the impact of various forms of model misfit on BMA predictions in simple situations points to prescriptive guidelines for “tuning” Zellner's <span><math><mi>g</mi></math></span> prior to obtain optimal predictions. The tuning of the prior distribution is obtained by considering theoretical properties that should be enjoyed by the optimal fits of the various models in the BMA ensemble. The methodology can be thought of as an “empirical Bayes” approach to modeling, as the data help to inform the specification of the prior in an attempt to attenuate the negative impact of influential cases.</p></div>","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"27 ","pages":"Pages 102-119"},"PeriodicalIF":1.9,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50178662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}